US20230128625A1 - Method for estimating object angle with high-angle analysis using large-scale mimo array antenna - Google Patents

Method for estimating object angle with high-angle analysis using large-scale mimo array antenna Download PDF

Info

Publication number
US20230128625A1
US20230128625A1 US17/728,381 US202217728381A US2023128625A1 US 20230128625 A1 US20230128625 A1 US 20230128625A1 US 202217728381 A US202217728381 A US 202217728381A US 2023128625 A1 US2023128625 A1 US 2023128625A1
Authority
US
United States
Prior art keywords
angle
matrix
pursuit
input signal
calculation model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
US17/728,381
Other languages
English (en)
Inventor
Hsiao-Ning WANG
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Cubtek Inc
Original Assignee
Cubtek Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Cubtek Inc filed Critical Cubtek Inc
Assigned to CUBTEK INC. reassignment CUBTEK INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: WANG, HSIAO-NING
Publication of US20230128625A1 publication Critical patent/US20230128625A1/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S13/06—Systems determining position data of a target
    • G01S13/42—Simultaneous measurement of distance and other co-ordinates
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
    • G01S2013/0236—Special technical features
    • G01S2013/0245—Radar with phased array antenna

Definitions

  • the present invention relates to estimation methods, and more particularly, to a method for estimating object angle with high-angle analysis using large-scale MIMO array antenna.
  • Antenna modules are widely used in modern society. For example, in the fields of wireless communication and radar detection, antenna modules are usually applied to transmit and receive wireless signals for signal transmission or position detection. In a multi-antenna signal source environment, it is desirable to resolve the issues of mutual interference among signals, the best signal transmission path, and selection of the optimal signal source in the technical field of antennas.
  • a car radar detection system is mainly applied for detecting objects around the vehicle, so as to confirm and recognize obstacles, thereby assisting the driver to avoid the risk of collision, or even facilitating an automatic driving function.
  • the detection principle of the radar system of vehicle is detecting the distance between the vehicle and the objects around the vehicle through the radar antenna during the moving process of the vehicle.
  • the solution of conventional radar works based on two to three transmitting antennas and three to four receiving antennas.
  • the imaging radar applies a plurality of multiple-input multiple-output (MIMO) antenna arrays for mapping the surrounding environment with high-resolution, thereby providing highly precise image quality without Doppler fuzzy sensing issues under any external environmental conditions and helping to achieve long-distance and wide-view decision making to avoid collisions between the vehicle and an object or another vehicle, so as to prevent traffic accidents.
  • MIMO multiple-input multiple-output
  • the target signal detection method used by the radar antenna of mobile phone can apply a signal detection method with longer calculation time for repeatedly searching the locations of the base stations and identifying the signal strength of the base stations, so as to receive a stronger signal.
  • the position of the base station is fixed, so that the position of the base station is easier to be identified.
  • the radar antenna of the vehicle is applied for preventing collisions during the movement of the vehicle, so that the vehicle radar antenna has to accurately detect the amounts and locations of objects in a short period of time. Therefore, a more accurate and efficient signal calculation method is needed for abovementioned objectives.
  • the present invention aims at resolving the issue of that the conventional target signal detection method is unable to accurately identify the position of the target object and the detection calculation time period thereof is relatively longer.
  • the present invention provides a method for estimating object angle with high-angle analysis using large-scale MIMO array antenna, which receives an input signal matrix through an array antenna, wherein the input signal matrix is a transmission signal or reflection signal of at least one target object.
  • the method for estimating object angle with high-angle analysis using large-scale MIMO array antenna comprises following steps: step S1: inputting an input signal matrix to a first calculation model, and inputting a plurality of angle related pursuit matrices for comparison; when the plurality of pursuit matrices satisfy conditions of the first calculation model, a target object amount and a rough object angle with respect to a location of a target object are obtained; S2: inputting the target object amount and the input signal matrix to a second calculation model for carrying out a singular value decomposition to obtain a noise matrix; S3: obtaining an iteration angle range from the rough object angle in S1; S4: inputting the plurality of pursuit matrices corresponding to the iteration angle range and the noise matrix to a third calculation model for an angle range iteration, thereby obtaining an accurate object angle.
  • the present invention accurately obtains information related to the target object amount and the angle thereof, and increases the calculation efficiency, allowing the present invention to be applied to various equipment.
  • FIG. 1 is a flow chart of the method for estimating object angle with high-angle analysis using large-scale MIMO array antenna in accordance with an embodiment of the present invention.
  • FIG. 2 is a detailed flow chart of step S1 in accordance with an embodiment of the present invention.
  • FIG. 3 is a schematic view of the operation status in accordance with the first embodiment of the present invention, illustrating that only one target object exists.
  • FIG. 4 is a schematic view of the operation status in accordance with the second embodiment of the present invention, illustrating that at least two target objects exist.
  • FIG. 5 is a curve diagram of the second embodiment of the present invention, illustrating that the target objects are located at a ⁇ 12-degree location and a 21-degree location, respectively.
  • FIG. 6 is a curve diagram of another embodiment of the present invention, illustrating that the target objects are located at a ⁇ 1-degree location and a 1-degree location, respectively.
  • Radar array antennas have various usages, such as object detection, weather observation, object tracking, and so on.
  • the radar array antenna used for detecting objects also applies different algorithms, such as orthogonal matching pursuit (OMP), MUltiple SIgnal Classification (MUSIC), and iterative sparse asymptotic minimum variance (SAMV) for signal identification and calculation.
  • OMP orthogonal matching pursuit
  • MUSIC MUltiple SIgnal Classification
  • SAMV iterative sparse asymptotic minimum variance
  • the object detection method of orthogonal matching pursuit first compares the correlation between the input signal matrix (reflection signal of the target object received by the radar array antenna) and angle-related steering vector, and then calculates residual value through the residue analysis in regression analysis according to the correlation between the input signal matrix and the steering vector, so as to obtain the amount and the positional angle of the object.
  • OMP has the property of fast calculation. However, to achieve the fast calculation, the detection accuracy of OMP is not optimal. In other words, OMP can only identify the target object in a certain angular range, instead of obtaining a precise angle information of the target object.
  • the object detection method of MUltiple SIgnal Classification first converts the reflection signal of the target object received by the radar array antenna into a covariance matrix, and carries out a Singular Value Decomposition (SVD) according to the amount of the target object and the covariance matrix to obtain a noise matrix, and finally carries out the iteration with the noise matrix and an angle related steering vector, thereby obtaining the angle information of the target object.
  • the MUSIC is able to obtain a highly accurate angle information.
  • the MUSIC has to deal with a huge amount of calculation, so that the calculation speed of MUSIC is relatively slow.
  • the MUSIC has to acquire the amount of the target object in advance before carrying out the calculation. Therefore, the MUSIC can only be applied in a relatively limited scope.
  • the method for estimating object angle with high-angle analysis using large-scale MIMO array antenna 100 receives an input signal matrix through an array antenna, wherein the input signal matrix is the transmission signal or reflection signal of at least one target object.
  • the array antenna is a multi-input multi-output (MIMO) array antenna, which uses multiple antennas on the transmission end to transmit a signal, respectively, and uses multiple antennas on the receiving end to receive and recover original information at the same time.
  • MIMO multi-input multi-output
  • the input signal matrix is obtained from the reflection signal which is reflected by the target object and received by the radar array antenna disposed on the vehicle, so as to be applied in an advanced driver assistance system (ADAS), for example but not limited to, being used for forward collision warning (FCW) or blind spot detection (BSD) functions.
  • ADAS advanced driver assistance system
  • FCW forward collision warning
  • BSD blind spot detection
  • the method of the present invention can be used for the detection of number, location, and angle of the target objects.
  • the present invention can also be applied for identifying connections between the mobile phone and the base stations.
  • connection signal of mobile phones the mobile phone needs to perform the connection detection with respect to a plurality of base stations to determine the optimal base station for connection, so that the method of the present invention is applicable as well.
  • the present invention comprises following steps:
  • Step S1 inputting the input signal matrix to a first calculation model (such as orthogonal matching pursuit, OMP), and inputting a plurality of angle related pursuit matrices for comparison.
  • a first calculation model such as orthogonal matching pursuit, OMP
  • the plurality of pursuit matrices satisfy conditions of the first calculation model, a target object amount and a rough object angle with respect to a location of a target object are obtained.
  • the plurality of pursuit matrices are formed of angle related steering vectors.
  • the steering vector is resented as formula (1)
  • v ⁇ ( k ) [ e - jksin ⁇ ⁇ 1 e - jk ⁇ 2 ⁇ sin ⁇ ⁇ 2 ⁇ e - jknsin ⁇ ⁇ n ] ( 1 )
  • v(k) is the steering vector.
  • a step S1a, step S1b, step S1c, step S1d, and step S1e are used to identify if the plurality of pursuit matrices satisfy the conditions of the first calculation model. Steps S1a, S1b, S1c, S1d, and S1e are illustrated below:
  • Step S1a the first calculation model comparing if there is a correlation between the input signal matrix and the plurality of pursuit matrices. If there is no correlation between one of the plurality of pursuit matrices and the input signal matrix, step S1a is repeated for carrying out the correlation comparison between the next plurality of pursuit matrices and the input signal matrix. If there is a correlation between one of the plurality of pursuit matrices and the input signal matrix, S1b is then performed.
  • the angle of the plurality of pursuit matrices ranges from ⁇ 90 to 90 degrees; the corresponding plurality of pursuit matrices are orderly compared with the input signal matrix for correlation comparison based on a first interval angle unit.
  • the correlation comparison compares the angular correlation between the plurality of pursuit matrices and the input signal matrix.
  • the value of the first interval angle ranges from 1/18 and 1/180 (1 degree to 10 degrees) of the angle range of the plurality of pursuit matrices. The user is allowed to set up the first interval angle according to actual demands.
  • Step S1b defining the pursuit matrix related to the input signal matrix as a first pursuit matrix, and performing a residual value calculation using the first pursuit matrix and the input signal matrix.
  • the first calculation model uses a least square method for residual value calculation.
  • Step S1c comparing the residual value with a residual threshold value. If the residual value presents a singular sharp reduction and is lower than or equal to the residual threshold value (indicating that the first pursuit matrix satisfies the conditions of the first calculation model), the first calculation model obtains the amount and the rough object angle of the target object. If the residual value is larger than the residual threshold value, the method proceeds to step S1d. Therein, if the residual value is larger than the residual threshold value, it is indicated that there may by a plurality of target objects, or that there is a correlation between the first pursuit matrix and the input signal matrix, but the correlation is minor.
  • Step S1d comparing the reduction degree of the residual value. If the reduction degree of the residual value is larger than a significant threshold value, it is indicated that there is a major correlation between the first pursuit matrix and the input signal matrix; also, the method proceeds to step S1e. If the reduction degree of the residual value is smaller than the significant threshold value, the method goes back to step S1a to repeat the correlation comparison between the input signal matrix and others of the plurality of pursuit matrices, until that the residual value above mentioned is smaller than or equal to the residual threshold value, whereby the first calculation model obtains the amount and the rough object angle of the target object. Therein, when the reduction degree of the residual value is smaller than the significant threshold value, it is indicated that there is a correlation between the first pursuit matrix and the input signal matrix, but the correlation is minor.
  • Step S1e storing the first pursuit matrix, and converting the input signal matrix into a second input signal matrix unrelated to the first pursuit matrix; also, the method goes back to step S1a to perform the correlation comparison between the second input signal matrix and others of the plurality of pursuit matrices, until that the residual value mentioned is smaller than or equal to the residual threshold value, whereby the first calculation model obtains the amount and the rough object angle of the target object.
  • Step S2 converting the input signal matrix into a covariance matrix, and inputting covariance matrix and the target object amount obtained from the first calculation model into a second calculation model for carrying out a singular value decomposition to obtain a noise matrix.
  • the input signal matrix is converted into the covariance matrix through formula (2):
  • R is the covariance matrix
  • y is the input signal matrix
  • S3 obtaining an iteration angle range from the rough object angle obtained in S1.
  • the iteration angle range is formed of the rough object angle plus and minus an allowance angle, and the allowance angle is smaller than the first interval angle.
  • the third calculation model carries out the iteration of the iteration angle range and the noise matrix through an orthogonal calculation method. After the iteration of the third calculation model, a curve of an angle and power is obtained, and the angle corresponding to the significant peak value on the curve is the accurate object angle.
  • a vehicle radar 1 is provided.
  • the vehicle radar 1 has a detection reference line L, which is arranged in perpendicular to the vehicle radar 1 .
  • a direction upward from the detection reference line L is defined as the negative direction
  • a direction downward from the detection reference line L is defined as the positive direction.
  • FIG. 3 which is in accordance with the first embodiment of the present invention, taking the detection reference line L as the basis, an object 2 exists at an angle of ⁇ 12 degrees with respect to the vehicle radar 1 .
  • FIG. 3 which is in accordance with the first embodiment of the present invention, taking the detection reference line L as the basis, an object 2 exists at an angle of ⁇ 12 degrees with respect to the vehicle radar 1 .
  • a first object 3 a and a second object 3 b exist, wherein the first object 3 a is located at ⁇ 12 degrees with respect to the vehicle radar 1 , and the second object 3 b is located at 21 degrees with respect to the vehicle radar 1 .
  • the actual detection steps are illustrated below:
  • step S1 the input signal matrix is inputted into the first calculation model, and the plurality of pursuit matrices are inputted for comparison.
  • the plurality of pursuit matrices satisfy the conditions of the first calculation model, the amount and the rough object angle of the target objects are obtained.
  • the input signal matrix is obtained from the reflection signal of the object 2 received by the vehicle radar 1 .
  • the plurality of pursuit matrices are formed of angle related steering vectors, and the range of the plurality of pursuit matrices ranges from ⁇ 90 to 90 degrees.
  • Step S1 proceeds through step S1a to step S1e for identifying if the plurality of pursuit matrices satisfy the conditions of the first calculation model.
  • step S1a the first calculation model orderly compares the correlation between the plurality of pursuit matrices and the input signal matrix based on the first interval angle unit from ⁇ 90 degrees to 90 degrees. When there is no correlation between one of the plurality of pursuit matrices and the input signal matrix, step S1a is repeated.
  • the first interval angle is 5 degrees.
  • the first calculation model compares the correlation between the pursuit matrix at ⁇ 40 degrees and the input signal matrix, based on the fact that the object 2 is at an angle ( ⁇ 12 degrees) which is significantly different from the angle ( ⁇ 40 degrees) of the pursuit matrix, the first calculation model determines that there is no correlation between the pursuit matrix at ⁇ 40 degrees and the input signal matrix. Then, the first calculation model repeats step S1a for performing the correlation comparison between the plurality of pursuit matrices at the next angle and the input signal matrix.
  • step S1b if there is a correlation between one of the plurality of pursuit matrices and the input signal matrix, step S1b is performed.
  • the first calculation model when the first calculation model performs the correlation comparison between the pursuit matrix at ⁇ 10 degrees and the input signal matrix, based on the fact that the object 2 is at an angle ( ⁇ 12 degrees) which is extremely close to the angle ( ⁇ 10 degrees) of the pursuit matrix, the first calculation model determines that there is a correlation between the pursuit matrix at ⁇ 10 degrees and the input signal matrix, and step S1b is then performed.
  • step S1b the pursuit matrix that are correlated to the input signal matrix are defined as the first pursuit matrix.
  • the first calculation model according to the first pursuit matrix and the input signal matrix, performs the least square method for calculating the residual value.
  • the first calculation model defines the pursuit matrix at ⁇ 10 degrees as the first pursuit matrix, and accordingly calculates the residual value. Therein, the initial value of the residual value is 100%.
  • step S1c the residual value is compared with the residual threshold value. If the residual value presents a singular sharp reduction and is lower than or equal to the residual threshold value, it is indicated that the first pursuit matrix satisfies the conditions of the first calculation model, and the first calculation model can obtain the amount and the rough object angle of the target object. If the residual value is larger than the residual threshold value, step S1d is performed.
  • the residual threshold value is set at 10%, and the calculated residual value is 5%. Because only one object 2 exists, and the locational angle ( ⁇ 10 degrees) of it is extremely close to the angle ( ⁇ 10) of the first pursuit matrix, the residual value presents a singular sharp reduction to be lower than the residual threshold value (5% ⁇ 10%). Therefore, the first calculation model obtains the rough object angle as ⁇ 10 degrees. Also, because the residual value is lower than the residual threshold value (5% ⁇ 10%) after a singular sharp reduction, the first calculation model obtains the fact that there is only one target object.
  • step S1c Based on the fact that, in the first embodiment, the amount (1) and the rough object angle ( ⁇ 10 degrees) of the target object are obtained in step S1c, the objective (obtaining the amount and the rough object angle of the target object) of step S1 is achieved, so that step S1d and step S1e are no longer needed to be carried out. The method proceeds to step S2.
  • step S2 the input signal matrix is converted into a covariance matrix.
  • the covariance matrix and the target object amount are inputted into the second calculation model for carrying out a singular value decomposition (SVD) to obtain a noise matrix.
  • the covariance matrix and the target object amount (1) are inputted into the second calculation model for carrying out a singular value decomposition to obtain the noise matrix.
  • step S3 the iteration angle range is obtained from the rough object angle plus and minus the allowance angle, wherein the allowance angle is smaller than the first interval angle.
  • the allowance angle is set to be 3 degrees. Therefore, the iteration angle range is from ⁇ 7 degrees to ⁇ 13 degrees based on the rough object angle ( ⁇ 10 degrees) plus and minus the allowance angle (3 degrees). Therein, the allowance angle (3 degrees) is smaller than the first interval angle (5 degrees).
  • step S4 the plurality of pursuit matrices corresponding to the iteration angle range and the noise matrix are inputted into a third calculation model, such that the third calculation model carries out an angle range iteration through an orthogonal calculation method, thereby obtaining the accurate object angle.
  • the plurality of pursuit matrices corresponding to the iteration angle range are orderly inputted into the third calculation model based on the second interval angle unit.
  • the second interval angle is set as 0.1 degrees.
  • the plurality of pursuit matrices corresponding to the iteration angle range ( ⁇ 7 to ⁇ 13 degrees) are orderly inputted based on the second interval angle (0.1 degree) unit together with the noise matrix into the third calculation model (such as MUltiple SIgnal Classification, MUSIC). Therefore, the third calculation model orderly carries out the angle iteration with the plurality of pursuit matrices corresponding to the iteration angle range ( ⁇ 7 to ⁇ 13 degrees) and the noise matrix through orthogonal calculation method, thereby obtaining the accurate target object angle ( ⁇ 12 degrees).
  • the third calculation model orderly carries out the angle iteration with the plurality of pursuit matrices corresponding to the iteration angle range ( ⁇ 7 to ⁇ 13 degrees) and the noise matrix through orthogonal calculation method, thereby obtaining the accurate target object angle ( ⁇ 12 degrees).
  • the curve of the angle and power is obtained. Because the locational angle of the object 2 is ⁇ 12 degrees, the curve would only contain one significant peak value at the position corresponding to ⁇ 12 degrees, indicating that the accurate angle is ⁇ 12 degrees.
  • the vehicle radar 1 detects that there is an object 2 existing at the locational angle of ⁇ 12 degrees through steps S1 to S4 of the present invention, and no other objects exist at other angles.
  • step S1 the input signal matrix is inputted into the first calculation model, and the plurality of pursuit matrices are inputted for comparison.
  • the plurality of pursuit matrices fulfill the conditions of the first calculation model, the amount and the rough object angle of the target object are obtained.
  • the input signal matrix is obtained from the reflection signal of the first object 3 a and the second object 3 b received by the vehicle radar 1 .
  • the plurality of pursuit matrices are formed of angle related steering vectors, and the range of the plurality of pursuit matrices ranges from ⁇ 90 to 90 degrees.
  • Step S1 proceeds through step S1a to step S1e to identify if the plurality of pursuit matrices fulfill the conditions of the first calculation model.
  • step S1a the first calculation model orderly compares the correlation between the plurality of pursuit matrices and the input signal matrix based on the first interval angle unit from ⁇ 90 degrees to 90 degrees. When there is no correlation between one of the plurality of pursuit matrices and the input signal matrix, step S1a is repeated.
  • the first interval angle is set at 5 degrees.
  • the first calculation model compares the correlation between the pursuit matrix at ⁇ 40 degrees and the input signal matrix, based on the fact that the locational angle of the first object 3 a ( ⁇ 12 degrees) and the locational angle of the second object 3 b (21 degrees) are significantly different from the angle ( ⁇ 40 degrees) of the pursuit matrix, the first calculation model determines that there is no correlation between the pursuit matrix at ⁇ 40 degrees and the input signal matrix. Then, the first calculation model repeats step S1a for performing the correlation comparison between the plurality of pursuit matrices at the next angle and the input signal matrix.
  • step S1b if there is a correlation between one of the plurality of pursuit matrices and the input signal matrix, step S1b is performed.
  • the first calculation model performs the correlation comparison between the pursuit matrix at ⁇ 15 degrees and the input signal matrix, based on the fact that the first object 3 a is at the angle ( ⁇ 12 degrees) which is slightly different from the angle ( ⁇ 15 degrees) of the pursuit matrix, the first calculation model determines that there is a correlation between the pursuit matrix at ⁇ 15 degrees and the input signal matrix, and step S1b is then performed.
  • step S1b the pursuit matrix that is correlated to the input signal matrix is defined as the first pursuit matrix.
  • the first calculation model according to the first pursuit matrix and the input signal matrix, performs the least square method for calculating the residual value.
  • the first calculation model defines the pursuit matrix at ⁇ 15 degrees as the first pursuit matrix, and accordingly calculates the residual value. Therein, the initial value of the residual value is 100%.
  • step S1c the residual value is compared with the residual threshold value. If the residual value presents a singular sharp reduction and is lower than or equal to the residual threshold value, it is indicated that the first pursuit matrix satisfies the conditions of the first calculation model, and the first calculation model can obtain the amount and the rough object angle of the target object. If the residual value is larger than the residual threshold value, step S1d is performed.
  • the residual threshold value is set at 10%, and the calculated residual value is 90%.
  • the residual value is still larger than the residual threshold value (90%>10%), so that the method has to proceed to the next step (step S1d).
  • step S1d the reduction degree of the residual value is compared. If the reduction degree of the residual value is smaller than the significant threshold value, the method goes back to step S1a to repeat the correlation comparison between the input signal matrix and others of the plurality of pursuit matrices, until that the residual value above mentioned is smaller than the residual threshold value. When the reduction degree of the residual value is larger than the significant threshold value, step S1e is performed.
  • the significant threshold value is set at 35%, and the calculated residual value is 90%.
  • the reduction degree (10% reduction) of the residual value is smaller than the significant threshold value (35%), meaning that there is a correlation between the first pursuit matrix (at ⁇ 15 degrees) and the input signal matrix, but the correlation is not major. Therefore, the first calculation model goes back to step S1a to perform the correlation comparison between the input signal matrix and the next pursuit matrix (at ⁇ 10 degrees).
  • the first calculation model repeats steps S1a to S1d.
  • step S1a because the locational angle ( ⁇ 12 degrees) of the first object 3 a is slightly different from the angle ( ⁇ 10 degrees) of the pursuit matrix, the first calculation model determines that there is a correlation between the pursuit matrix at ⁇ 10 degrees and the input signal matrix.
  • step S1b the pursuit matrix at the angle of ⁇ 10 degrees is defined as the first pursuit matrix, and the residual value is accordingly calculated. Meanwhile, the initial value of the residual value is 90%.
  • the calculated residual value is 50%, and the residual value is still larger than the residual threshold value (50%>10%). Therefore, the method has to proceed to step S1d.
  • step S1d the reduction degree of the residual value (reduced by 40% from 90% to 50%) is larger than the significant threshold value (35%), so that the method proceeds to step S1e.
  • the reduction degree (40%) of the residual value is larger than the significant threshold value, meaning that there is a major correlation between the first pursuit matrix (at the angle of ⁇ 10 degrees) and the input signal matrix.
  • step S1e the first pursuit matrix is stored, and the input signal matrix is converted into the second input signal matrix which is unrelated to the first pursuit function; also, the method goes back to step S1a to perform the correlation comparison between the second input signal matrix and others of the plurality of pursuit matrices, until that the residual value is lower than or equal to the residual threshold value, whereby the first calculation model obtains the amount and the rough object angle of the target object.
  • the first pursuit matrix (at the angle of ⁇ 10 degrees) is stored, and the input signal matrix is converted into the second input signal matrix (unrelated to the angle of ⁇ 10 degrees) which is unrelated to the first pursuit matrix (at ⁇ 10 degrees). Then, the method goes back to step S1a to perform the correlation comparison between the second input signal matrix and the next pursuit matrix (at the angle of ⁇ 5 degrees), until that the residual value is lower than or equal to the residual threshold value (10%). Therein, the first pursuit matrix (at ⁇ 10 degrees) is stored, meaning that one of the rough object angles is ⁇ 10 degrees.
  • the first calculation model determines that, through steps S1a to S1e, there are two target objects (first object 3 a and second object 3 b ), and the rough object angle of the objects is ⁇ 10 degrees and 20 degrees, respectively.
  • step S2 the input signal matrix is converted into the covariance matrix.
  • the covariance matrix and the target object amount are inputted into the second calculation model for singular value decomposition (SVD) to obtain a noise matrix.
  • the covariance matrix and the target object amount (2) are inputted into the second calculation model for singular value decomposition to obtain the noise matrix.
  • step S3 the iteration angle range is obtained from the rough object angle plus and minus the allowance angle, wherein the allowance angle is smaller than the first interval angle.
  • the allowance angle is set to be 3 degrees. Therefore, the iteration angle ranges are from ⁇ 7 degrees to ⁇ 13 degrees and 17 degrees to 23 degrees based on the rough object angles ( ⁇ 10 degrees and 20 degrees) plus and minus the allowance angle (3 degrees). Therein, the allowance angle (3 degrees) is smaller than the first interval angle (5 degrees).
  • step S4 the plurality of pursuit matrices corresponding to the iteration angle range and the noise matrix are inputted into a third calculation model, such that the third calculation model carries out the angle iteration through the orthogonal calculation method, thereby obtaining the accurate object angle.
  • the plurality of pursuit matrices corresponding to the iteration angle range are orderly inputted into the third calculation model based on the second interval angle unit.
  • the second interval angle is set as 0.1 degrees.
  • the plurality of pursuit matrices corresponding to the iteration angle range ( ⁇ 7 to ⁇ 13 degrees and 17 to 23 degrees) are orderly inputted based on the second interval angle (0.1 degrees) unit together with the noise matrix into the third calculation model. Therefore, the third calculation model orderly carries out the angle iteration with the plurality of pursuit matrices corresponding to the iteration angle range ( ⁇ 7 to ⁇ 13 degrees and 17 to 23 degrees) and the noise matrix through the orthogonal calculation method, thereby obtaining the accurate target object angle ( ⁇ 12 degrees and 21 degrees).
  • the curve of the angle and power is obtained. Because the locational angle of the first object 3 a is ⁇ 12 degrees and the locational angle of the second object 3 b is 21 degrees, the curve would only have significant peak values at the positions corresponding to ⁇ 12 degrees and 21 degrees, indicating that the accurate angles are ⁇ 12 degrees and 21 degrees.
  • the vehicle radar 1 detects that, through steps S1 to S4, there are a first object 3 a located at the angle of ⁇ 12 degrees and a second object 3 b located at the angle of 21 degrees, with no other objects existing at other angles.
  • the obtained curve contains significant peak values at the positions of ⁇ 1 degree and 1 degree, indicating that there is a target object at the ⁇ 1 degree and 1 degree, respectively. More particularly, as shown by FIG. 6 , it is clear that even if different target objects are extremely close to each other (at the angle of ⁇ 1 degree and 1 degree), the curve is still able to accurately present significant peak values at the position of corresponding angles ( ⁇ 1 degree and 1 degree). Therefore, the user is allowed to efficiently and clearly know that there is a target object at the location of ⁇ 1 degree and 1 degree, respectively through the curve.
  • the present invention carries out the calculation on a larger scale first through the first calculation model, such that most angles unrelated to the input signal matrix are excluded first, so as to obtain the target object amount and the rough object angle, increasing the calculation speed of the present invention.
  • the present invention through the combination of the second calculation model and the third calculation model, break down the rough object angle and obtain the accurate angle through angle iteration. Therefore, the present invention identifies the target object location with higher accuracy.
  • the curve obtained by the third calculation model of the present invention only shows significant peak value at the position corresponding to the accurate object angle. Therefore, the user is allowed to clearly know the angle information of the target object.

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Radar Systems Or Details Thereof (AREA)
  • Variable-Direction Aerials And Aerial Arrays (AREA)
US17/728,381 2021-10-27 2022-04-25 Method for estimating object angle with high-angle analysis using large-scale mimo array antenna Pending US20230128625A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
TW110139954 2021-10-27
TW110139954A TWI813046B (zh) 2021-10-27 2021-10-27 利用大型多輸入多輸出陣列天線進行高角度解析之目標物角度估測方法

Publications (1)

Publication Number Publication Date
US20230128625A1 true US20230128625A1 (en) 2023-04-27

Family

ID=81323555

Family Applications (1)

Application Number Title Priority Date Filing Date
US17/728,381 Pending US20230128625A1 (en) 2021-10-27 2022-04-25 Method for estimating object angle with high-angle analysis using large-scale mimo array antenna

Country Status (4)

Country Link
US (1) US20230128625A1 (zh)
CN (1) CN116027330A (zh)
DE (1) DE102022117030A1 (zh)
TW (1) TWI813046B (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116908834B (zh) * 2023-09-06 2023-11-24 齐鲁工业大学(山东省科学院) 基于毫米波雷达的多目标定位跟踪方法及系统

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI470257B (zh) * 2013-10-07 2015-01-21 Univ Nat Chiao Tung 角度估計檢測方法及電子裝置
CN103760546B (zh) * 2014-01-23 2015-11-18 西安电子科技大学 一种雷达用低空目标波达方向估计方法
US10505620B2 (en) * 2016-04-12 2019-12-10 Mitsubishi Electric Corporation Receiving apparatus and receiving method, and program and recording medium
TWI634342B (zh) * 2017-08-30 2018-09-01 啟碁科技股份有限公司 角度估測方法及雷達系統
CN108828504B (zh) * 2018-03-30 2022-06-10 西安电子科技大学 基于部分相关波形的mimo雷达目标方向快速估计方法
DE102018210070A1 (de) * 2018-06-21 2019-12-24 Robert Bosch Gmbh Verfahren zur Kalibrierung eines MIMO-Radarsensors für Kraftfahrzeuge
TWI693419B (zh) * 2019-02-13 2020-05-11 國立交通大學 訊號處理方法
CN111220977B (zh) * 2020-01-16 2022-04-08 深圳大学 基于角度及频率域滤波的似然music低仰角估计方法
CN112731326A (zh) * 2020-12-24 2021-04-30 深圳大学 非高斯噪声下的雷达信号波形与目标角度联合估计方法
CN113093136B (zh) * 2021-03-31 2022-06-10 桂林电子科技大学 一种频率分集阵列雷达去目标位置模糊成像方法

Also Published As

Publication number Publication date
DE102022117030A1 (de) 2023-04-27
TW202206844A (zh) 2022-02-16
CN116027330A (zh) 2023-04-28
TWI813046B (zh) 2023-08-21

Similar Documents

Publication Publication Date Title
US20210223376A1 (en) Techniques for Cooperative Passive Positioning
Dias et al. Position and LIDAR-aided mmWave beam selection using deep learning
US20230128625A1 (en) Method for estimating object angle with high-angle analysis using large-scale mimo array antenna
Radmard et al. Antenna placement and power allocation optimization in MIMO detection
US10805022B2 (en) Method of using time domain subspace signals and spatial domain subspace signals for location approximation through orthogonal frequency-division multiplexing
US20220035019A1 (en) Signal processing method, radar system, and vehicle
US20110285571A1 (en) Sensor and alignment adjusting method
JPWO2013140747A1 (ja) レーダ装置
US20220326376A1 (en) Signal processing method and apparatus
CN109991596A (zh) 用于检测目标的设备和方法
Reus-Muns et al. Deep learning on visual and location data for V2I mmWave beamforming
CN114966589A (zh) 一种基于毫米波雷达的多目标检测方法
Moayyed et al. Clustering algorithms and validation indices for mmWave radio multipath propagation
US20230127933A1 (en) High angular resolution processing method for mimo system
CN112098965B (zh) 一种基于毫米波雷达的载体姿态角解算方法
CN113325363A (zh) 确定波达方向的方法、装置及相关设备
EP4116738A1 (en) Radar device and method for classifying objects
JP2021124295A (ja) 物標判別装置、レーダ装置、および物標判別方法
JPWO2017104224A1 (ja) 運転支援情報伝送システム、受信機、運転支援システム、及び運転支援情報伝送方法
CN113115437B (zh) 定位标签与基站前后位置关系判定方法及系统
CN116260497A (zh) 波束选取的方法、装置、电子设备及介质
Ortiz et al. Vehicle tracking through vision-millimeter wave Doppler shift fusion
CN114884585A (zh) 多维联合决策的转发式干扰感知与防护方法及装置
US20230288557A1 (en) Electronic device, method for controlling electronic device, and program
Brambilla et al. Multi-Sensor Assisted Cooperative Beam Tracking for mmWave Vehicle-to-Vehicle Communication

Legal Events

Date Code Title Description
AS Assignment

Owner name: CUBTEK INC., TAIWAN

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:WANG, HSIAO-NING;REEL/FRAME:059700/0189

Effective date: 20220413

STPP Information on status: patent application and granting procedure in general

Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION